-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
1/19
Contributors to Uncertainty
Representative
and ApplicableQC Data
Transportation/Storage/Handling:
shipping time, container and temperature NA / (FS)
lab storage time, conditions and temperature NA / (LCS, FS)
contamination in lab storage areas NA / LCS, FB
Laboratory Subsampling
sample nonhomogeneity DUPblending techniques DUP
sample size DUPSample Preparation:
volumetric glassware LCS, DUPdispensing device LCS, DUP
balance LCS, DUPtemperature LCS, DUP
sample extraction LCS, DUPextractant background LCS, DUP, MB
Lab Environmental Conditions:
temperature variance LCS, DUPhumidity variance LCS, DUP
Analysts:
different analysts LCS, DUPanalyst training level and experience
LCS, DUP
data interpretation by analyst LCS, DUPMeasuring
Instruments:
instrument stability LCS, DUP
carry over effects LCS, DUP
Common Contributors to M(representative list - ma
(QC sample types in this list are typical of those u
addition
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
2/19
day to day calibration differences LCSinterferences LCS
Calibration Standards/Reference
Materials:
preparation variances LCS
calibration stock material uncertainty CERTIFICATE
LCS reference material uncertainty CERTIFICATETest Procedure
Variations:
variation within and between reagent lots LCSextraction or
digestion times, temperatures,
and conditions LCSsample dependent modifications LCS
desorption efficiencies within and between
lots for sorbents LCSData Manipulation:
sampling media/blank correction LCS, MBinstrument blank
correction LCS
accuracy of calculations LCS
area or air volume sampled NA
Where:
DUP = Duplicate, resulting from sub-sampling of a bul
NA = Not Applicable
FB = Field Blank
FS = Field Spike - Not typically conducted unless part
is responsible for field sampling.
LCS = Laboratory Control Standard, matrix matched a
batch
MB = Method or Matrix Blank
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
3/19
Comments to Clarify Contributor Effects
NOTE: This is not part of analytical uncertainty, but must be
considered by labs
providing sampling and when providing guidance regarding sample
packaging and
shipping. Usually no impact if recommended shipping conditions
and holding times in
referenced methods are maintained. Improper packing materials,
bulks shipped w/samples,
etc. may adversely affect data. Field blanks, field spiked
samples or duplicate field samples
shipped with samples or included in method validation studies
may reflect these contributors.
Field variability (FS/DUP) is only considered when lab is
responsible for sampling.
Usually no impact if recommended storage conditions and holding
times in reference
methods are maintained.Impact is monitored per sample batch only
if LCS samples are
prepared on receipt and stored with field samples. Field spiked
samples or stability study
samples included in some method validation studies may also
reflect these contributors. Field
variability (FS/DUP) is only considered when lab is responsible
for sampling.
Usually no impact if recommended storage conditions and holding
times in reference
methods are maintained. Improper storage such as sorbent tubes
stored with bulk solvent
samples or near solvent sources may adversely affect data.
Impact is monitored only if LCS
is prepared on receipt & stored with field samples. Field
blank can be used to assess
contamination from collection, transport, and storage.
NOTE:This primarily applies to bulk/solid samples which requires
use of laboratory duplicate
sample data to determine impact on uncertainty. It is not
applicable for LCS/LCSD media
spike duplicates.
Sample composition, etc.
Stirring, sieving, grinding, etc.
Large enough to allow adequate subsampling.
NA for Class A; applies for graduated tubes or cylinders,
etc.
Pipettes, and other types of dispensers that are not Class
A.
Balance error is often insignificant compared to other MU
sources.
Hot plate or ashing temperatures.
Applies to LCS or DUP if it goes throughthe entire sample
preparation process.
Analyte or interferant is present in acids, solvents, etc.
Room temp during bulk asbestos, gravimetry, etc. processes.
Gravimetry involving hydroscopic media, etc.
Must use inter-analyst instead of intra-analyst repeat data,
where applicable.
Must use inter-analyst instead of intra-analyst repeat data,
where applicable.
Chromatographic peak ID, interference corrections, etc. Must use
inter-analyst instead of
intra-analyst repeat data, where applicable.
Baseline drift, repeatability of averaged readings, lab
environmental stability, etc.
Impact of high samples on following sample readings; can be
monitored by proper use of
CCBs.
asurement Uncertainty Chemical Analyses not be all inclusive for
all types of analyses)
tilized in AIHA-LAP, LLC laboratories) See
l tabbed sheets for examples
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
4/19
Variation in instrument response and calibration process
Due to matrix, inter-element effects, co-eluting GC peaks,
etc.
Due to analysts, balances, dispensing devices used, etc
Obtain from certificate or estimate, can be ignored if less than
1/3 of the largest contributor.Only has impact when LCS data are
used to correct customer sample results. Obtain from
certificate or estimate. Can be ignored if less than 1/3 of the
largest contributor. Note that
use of an LCS with a large uncertainty can result in over
estimation of overall analytical
uncertainty.
Similar to extractant background effects under Sample
Preparation above.
May affect complete dissolution of analyte or loss of material
in some cases.
Changes in conditions due to sample size, customer requests,
etc.
May vary by lot or manufacturer; also applies to diffusion rates
for passive monitors.
When significant and when data are blank corrected.
When allowed.
Manual, spreadsheet, LIMS, etc.
Typically provided by the customer. This is not part of
analytical uncertainty, but must be
considered by labs providing sampling and providing combined
sampling and analytical
uncertainty.
k (NOTE: NOT LCS/LCSD duplicate spiked sampling media)
of sampling method validation. Should be considered only when
laboratory
nd typically taken through the entire analytical process with
each sample
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
5/19
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
6/19
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
7/19
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
8/19
Contributors to UncertaintyRepresentative andApplicable QC Data
Comments to Clarify Contributor Effects
Transportation/Storage/Handling
shipping time, container & temperature NA Limited impact on
most sorbent tubes
lab storage time, conditions & temperature NA
Usually no impact if recommended storage conditions and hold
times are
samples are representative if prepared on receipt & stored
with field sampl
impact if recommended storage conditions and holding times are
maintain
samples or stability study samples included in some method
validation stud
these contributors.
contamination in lab storage areas NAUsually no impact if
appropriate storage conditions are maintained. Field
assess contamination from collection, transport, and storage
Laboratory Subsampling sample nonhomogeneity NA Not applicable
to sorbent tube analysis
blending techniques NA Not applicable to sorbent tube
analysis
sample size NA Not applicable to sorbent tube analysis
Sample Preparation:
volumetric glassware LCS Same type of glassware used for samples
and LCS
dispensing device LCS Same type of dispensing device
balance NA Not applicable to sorbent tube analysis
temperature NA Not applicable to sorbent tube analysis
sample extraction LCS Applies to LCS if goes through sample
preparation
extractant background LCS, MB Analyte or interferant in solvents
or other prep reagents used, etc.
Lab Environmental Conditions:
temperature variance LCS LCS results reflect any temperature
effects on chromatography instrumenhumidity variance NA Not
applicable to sorbent tube analysis
Analysts:
different analysts LCS LCS results reflect variability due to
different analysts, as applicable, on dif
analyst training level & experience LCS LCS results reflect
variability due to different analysts, as applicable, on dif
data interpretation by analyst LCS LCS results reflect
variability due to different analysts, as applicable, on dif
Measuring Instruments:
instrument stability LCS LCS results reflect instrument
variability on different days
carry over effects LCS LCS results reflect instrument
variability on different days
day to day calibration differences LCS LCS results reflect
instrument variability on different days
interferences LCS LCS results reflect instrument variability on
different days
Calibration Standards/Reference Materials:
preparation variances LCS Due to analysts, dispensing devices
used, etc
calibration stock material uncertainty CERTIFICATE Obtain from
certificate or estimate
LCS reference material uncertainty CERTIFICATEUse if customer
sample data corrected for desorption efficiency. Obtain fr
estimate.
Test Procedure Variations
variation within and between reagent lots LCS LCS subjected to
same treatment as customer samples
extraction or digestion times and temps
LCSLCS subjected to same treatment as customer samples
sample dependent modifications LCS LCS subjected to same
treatment as customer samples
desorption efficiencies within and between lots
for sorbent tubes LCS LCS subjected to same treatment as
customer samples
Chemical Analyses of Sorbent Tubes using Chromatography
See Example Calculations (to the right of the table)
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
9/19
for sorbent tubes LCS LCS subjected to same treatment as
customer samples
LCS = Laboratory Control Standard, typically taken through the
entire analytical process with each sample batch
MB = Method or matrix blank
NA = Not Applicable
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
10/19
LCS mg LCS %Rec LCSD mg
LCSD
%Rec LCS mg LCS %Rec LCSD mg LCSD %Rec
0.0876 101.0 0.0884 102.0 0.0862 106.4 0.0832 96.0
0.0824 95.0 0.0850 98.0 0.0799 98.6 0.0812 93.7
0.0867 100.0 0.0858 99.0 0.0839 103.6 0.0825 95.2
0.0832 96.0 0.0841 97.0 0.0923 114.0 0.0874 100.8
0.0789 91.0 0.0789 91.0 0.0847 104.6 0.0819 94.5
0.0798 92.0 0.0806 93.0 0.0702 86.7 0.0907 104.6
0.0850 98.0 0.0850 98.0 0.0632 78.0 0.0676 78.0
0.0824 95.0 0.0867 100.0 0.0933 115.2 0.0870 100.30.0858 99.0
0.0867 100.0 0.1005 124.1 0.0924 106.6
0.0850 98.0 0.0850 98.0 0.0987 121.9 0.0933 107.6
0.0876 101.0 0.0858 99.0 0.0826 102.0 0.0919 106.0
0.0832 96.0 0.0884 102.0 0.0913 112.7 0.0756 87.2
0.0884 102.0 0.0919 106.0 0.0770 95.1 0.0920 106.1
0.0867 100.0 0.0893 103.0 0.0910 112.3 0.0960 110.7
0.0902 104.0 0.0884 102.0 0.0832 102.7 0.0928 107.0
0.0850 98.0 0.0876 101.0 0.0984 121.5 0.0884 102.0
0.0876 101.0 0.0902 104.0 0.0872 107.7 0.0861 99.3
0.0834 96.2 0.0892 102.9 0.0809 99.9 0.0655 75.5
0.0819 94.5 0.0812 93.7 0.0926 114.3 0.0956 110.3
0.0797 91.9 0.0784 90.4 0.0979 120.9 0.0889 102.5
98.2 103.1
4.0 11.5
RSD 4.05% RSD 11.2%
8.1% % 22.3% %
Bias @ 103.1% Rec = 3.1%
No significant background in method blank. No significant
background in method blank.
40 point Mean % Rec 40 point Mean % Rec
Example of reporting for 0.085 mg n=butanol: 0.0butanol with an
analytical uncertainty of +/- 0.019
95% confidence level with probable bias of 0.003
Reference material used for calibration indicates assay
of 99.5%. Expanded uncertainty of reference material
estimated at 0.5% (95%) Divide by 2 to yield 0.25 % Rel.
SD. Insignificant compared to 4.0% can be eliminated
from calculation
Reference materials used for calibration and LCS
preparation indicate assay of 99.5%. Expanded un
of reference materials estimated at 0.5% (95%). Di
to yield 0.25 % Rel. SD. Insignificant compared to
can be eliminated from calculation
Example of reporting for 0.085 mg toluene:0.085 mg toluene with
an analytical uncertainty of +/-
0.007 mg at the 95% confidence level with probable bias
of -0.002 mg
mg Toluene on Charcoal Tube using CS2Desorption per NIOSH
1501
LCS/LCSD = 0.0867 mg toluene spike/tube
Expanded MU @
95% Conf k=2
Bias @ 98.2% Rec = -1.8%
Examples of Analytical Measurement Uncertainty for Analysis of
Organic Solvents on Charcoal Tubes
mg n-Butanol on Charcoal Tube using CS2Desorption per NIOSH
1401Mod
LCS/LCSD = 0.0810 mg n-butanol spike /tube
(includes correction for DE = 0.80)
40 point Std Dev 40 point Std Dev
Expanded MU @
95% Conf k=2
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
11/19
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
12/19
5 mg n- g at the
g
certainty
ide by 2
11.2%
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
13/19
Contributors to Uncertainty
Representative
and Applicable
QC Data
Transportation/Storage/Handling
shipping time, container & temperature NAlab storage time,
conditions & temperature NA
contamination in lab storage areas NALaboratory Subsampling
sample nonhomogeneity DUPblending techniques DUP
sample size DUPSample Preparation:
volumetric glassware LCS, DUPdispensing device LCS, DUP
balance LCS, DUPtemperature LCS, DUP
sample extraction LCS, DUPextractant background LCS, DUP, MB
Lab Environmental Conditions:
temperature variance NAhumidity variance NA
Analysts:
different analysts LCS, DUPanalyst training level &
experience LCS, DUP
data interpretation by analyst LCS, DUPMeasuring
Instruments:
instrument stability LCS
carry over effects LCS, DUPday to day calibration differences
LCS
interferences DUP, MS
Calibration Standards/Reference Materials:
preparation variances LCS, DUP
calibration stock material uncertainty CERTIFICATELCS reference
material uncertainty NA
Test Procedure Variations
variation within and between reagent lots LCS
extraction or digestion times and temps LCS
sample dependent modifications LCS
Example of Contributor
Chemical Analyses of L
See Example Calculat
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
14/19
desorption efficiencies within and between lots for
sorbent tubes NAData Manipulation:
sampling media blank correction NAinstrument blank
correction
LCS
Accuracy of calculations LCS
DUP = Duplicate, resulting from sub-sampling of a bulk (N
MB = Method or matrix blankNA = Not Applicable
FB = Field Blank
FS = Field Spike
LCS = Laboratory Control Standard, matrix matched and t
sample batch
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
15/19
Comments to Clarify Contributor Effects
No impact on bulk paint samples from transportation, storage or
normal handling
LCS
mg/Kg
True
value
mg/Kg
195.4 208.8
189.6 208.8
4157 4490Sample composition, etc. 4186 4490Stirring, sieving,
grinding, etc 184 208.8Large enough to allow adequate subsampling
184 208.8
4641 4490NA for Class A; applies for graduated tubes or
cylinders, etc. 4831 4490pipettes, and other types of dispensers
not Class A 179 208.8balance error is often insignificant compared
to other MU sources 177 208.8Hot plate or ashing temperatures 4539
4490
Applies to LCS or DUP if goes through sample preparation 4858
4490Analyte or interferant in acids, or other reagents 198
208.8
207 208.8No impact on bulk paint samples 4458 4490No impact on
bulk paint samples 4514 4490
Analyst contributors affect all aspects of analysis from
subsampling through data
manipulation 184 208.8
187 208.8
4551 4490
4580 4490
179 208.8Baseline drift, repeatability of averaged readings, etc
186 208.8Impact of high samples on following sample readings; can
be monitored by proper
use of CCBs 4245 4490
4302 4490Due to matrix, inter-element effects, etc. Cannot be
routinely determined for
typical industrial hygiene sampling media 188 208.8
202 208.8Due to analysts, balances, dispensing devices used, etc
4217 4490
Obtain from certificate or estimate 4300 4490Sample results not
corrected for LCS recovery 203 208.8
190 208.8
Similar to extractant background effects under Sample
Preparation above 30 point Mean % Rec.
May affect complete dissolution of analyte or loss of material
in some cases
Changes in conditions due to sample size, customer requests, etc
RSD
30 point Std Dev
s to Measurement Uncertainty
ad (Pb) in Paint using ICP-AES
ons (to the right of the table)
Examp
Pb in Paint using hotblock acid
Analysis in accordance with EP
(Mod)
LCS Recovery of Paint SRM 258
or SRM 2581 at 4490 +/
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
16/19
No sampling media with bulk samples
when allowed
Manual, spreadsheet, LIMS, etc
OTE: NOT LCS/LCSD duplicate spiked sampling media)
ypically taken through the entire analytical process, with
each
Example o
4400 mg/K
confidenc
Example a
Expanded
Bias = 440
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
17/19
LCS % Rec
mg/Kg
Sample
mg/Kg
Sple Dup
Std Dev
(S) CV CV2
93.6 1604 1502 72.12 0.0464 0.0022
90.8 511 602 64.35 0.1156 0.0134
92.6 9470 8794 478.00 0.0523 0.0027
93.2 161 93 48.08 0.3786 0.1433
88.1 1683 1411 192.33 0.1243 0.0155
88.1 956 830 89.10 0.0998 0.0100
103.4 23470 26570 2192.03 0.0876 0.0077
107.6 625 730 74.25 0.1096 0.0120
85.7 723 472 177.48 0.2970 0.0882
84.8 23000 22000 707.11 0.0314 0.0010
101.1 14190 13900 205.06 0.0146 0.0002
108.2 10350 9142 854.18 0.0876 0.0077
94.8 5702 5854 107.48 0.0186 0.0003
99.1 109 122 9.19 0.0796 0.0063
99.3 7079 6427 461.03 0.0683 0.0047
100.5 196 186 7.07 0.0370 0.0014
88.1 14510 14300 148.49 0.0103 0.0001
89.6 15710 17150 1018.23 0.0620 0.0038
101.4 1187 1192 3.54 0.0030 0.0000
102.0 9265 9246 13.44 0.0015 0.0000
85.7 4240 3918 227.69 0.0558 0.0031
89.1 1979 1574 286.38 0.1612 0.0260
94.5 2357 3068 502.75 0.1853 0.0344
95.8 2254 3062 571.34 0.2150 0.0462
90.0 53160 44300 6264.97 0.1286 0.0165
96.7 24810 23000 1279.86 0.0535 0.0029
93.9 22860 22930 49.50 0.0022 0.0000
95.8 1133 1125 5.66 0.0050 0.000097.2 151 96 38.89 0.3149
0.0992
91.0 7774 4366 2409.82 0.3970 0.1576
94.7 CV 0.7064
6.3 CV pooled = (CV2/30) = 0.1534 15.3%
6.7%
Sample duplicata data in mg/Kg for Pb in Paint using
hotblock acid digestion and ICP-AES in accordance with
EPA SW-846 3050 & 6010 (Mod)
les of Analytical Measurement Uncertainty for Lead in Paint
digestion and ICP-AES
SW846 3050 & 6010
2 at 208.8 +/- 4.9 mg/Kg
110 mg/Kg Pb
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
18/19
Combined Rel. Std Dev (SDc) = [ SD12+ SD22]
SDc = [(6.7)2+ (15.3)2] = 16.7%
33.4%
f reporting for 4400 mg/Kg Pb in paint:
g Pb in paint with an analytical uncertainty of +/- 1500 mg/Kg
at the 95%
level and a probable bias of -230 mg/Kg
nalytical uncertainty for 4400 mg/Kg Pb in paint sample:
analytical uncertainty of 4400 mg/Kg Pb in paint = 4400 X0.334=
1470 mg/kg
mg/kg X -0.053 = 233.2 mg/kg
Bias @ 94.7% Rec of LCS = -5.3%
Reference material used for calibration indicates concentration
and expanded
uncertainty of 1001 +/- 5 ug/mL at 95% confidence level.
Expanded uncertainty
divided by 2 to yield 0.25 % Rel. SD. Insignificant compared to
15.2% and can
be eliminated from calculation. No significant background in
method blank.
Expanded MU @ 95% Conf (k=2) =
-
5/21/2018 Example Chemistry Measurement Uncertainty Calculations
V3 Corrected 031312
19/19
RSD